Turn a polars series to a dataframe

to_frame

In Polars, a series is a one-column table, while a dataframe is a multi-column table. However, it’s possible to convert a series into a dataframe. Below is a series of cities.

shape: (5_339,)

City
str
"GANADO"
"WEST MEMPHIS"
"MIAMI"
"BRADENTON"
"BOISE"
…
"SHELBYVILLE"
"DAYTON"
"WICHITA FALLS"
"GULF BREEZE"
"SWEET SPRINGS"

Convert series to dataframe

Notice that the series above has a shape of (5339,). This indicates that it contains 5339 values. Polars doesn’t need to display the number of columns in the shape because a series always has one column, which is already named City.

To convert the series to a dataframe, use the to_frame method like this:

(ser
 .to_frame()
 )

shape: (5_339, 1)

City
str
"GANADO"
"WEST MEMPHIS"
"MIAMI"
"BRADENTON"
"BOISE"
…
"SHELBYVILLE"
"DAYTON"
"WICHITA FALLS"
"GULF BREEZE"
"SWEET SPRINGS"

Why convert to dataframe

You may want to save your series data to a CSV file, but calling ser.write_csv() will result in an attribute error because write_csv is not a method of a series; it belongs to dataframes. So, to save your series data, you must first convert it to a dataframe and then save it.

Start learning Polars with my Polars course.